List your company

Reasoning Fidelity on AIME 2024 (ref6)

91.7Top-1 AccCASK
85.87687.38888.990.412Apr 13, 2026

Evaluation results

MethodLinks
CASKBudget=5122026.0491.799.90.2338.743.5
CASKBudget=3842026.0490.799.70.2688.761.6
TriAttentionBudget=5122026.0489.499.10.3338.743.6
CASKBudget=2562026.0488.499.20.3598.765.3
TriAttentionBudget=3842026.0488.298.90.3838.761.6
TriAttentionBudget=2562026.0486.1980.4638.765.3